There is a shift underway in B2B commerce that most leaders have not fully absorbed. It is not about making human buyers more efficient with AI tools. It is about the buyer being replaced, in part, by an autonomous system that researches, evaluates, negotiates, and purchases on behalf of the organisation, often without a human ever entering the conversation.
This is agentic commerce: the rise of AI agents as active participants in the procurement cycle. And it is arriving faster than most go-to-market strategies have accounted for.
The Evidence from Research
Gartner’s Top Strategic Predictions for 2026 and Beyond put the trajectory in stark terms: by 2028, 90% of all B2B buying will be AI-agent intermediated, channeling more than $15 trillion in spend through AI-driven exchanges. These are not future-looking projections built on optimism; they are extrapolations from adoption already underway. Gartner’s own research notes that new commercial models will emerge featuring high-frequency, frictionless sales powered by AI agents that can radically compress the sales cycle across a wide range of business and technology purchases.
Forrester’s data reinforces the near-term reality. In its B2B Marketing, Sales and Product Predictions 2026, Forrester found that 61% of purchase influencers say their organisation has or will deploy a private generative AI engine to support purchasing. The immediate implication: by 2026, at least one in five B2B sellers will be compelled to respond to AI-powered buyer agents issuing dynamically generated counteroffers; not human procurement managers deliberating over terms, but automated systems negotiating on behalf of organisations in real time.
6 Numbers B2B Leaders Need to Know
- 90% of B2B buying will be AI-agent intermediated by 2028
- $15T in B2B spend will flow through AI agent exchanges by 2028
- 61% of purchase influencers already use or plan to use a private AI engine for buying
- 1 in 5 sellers will face AI buyer agents negotiating deals autonomously in 2026
- 22 stakeholders in a typical enterprise purchase — 13 internal, 9 external
- $10B+ in enterprise value at risk in 2026 from ungoverned AI in go-to-market
The buying group itself is simultaneously growing and changing shape. Forrester’s State of Business Buying 2026 found that the typical enterprise purchasing decision now involves 13 internal stakeholders and nine external influencers. Procurement professionals are decision-makers in 53% of buying cycles, engaging from the very start of the process. These stakeholders scrutinise features and functions to assess efficiency and productivity, interacting more frequently with sales representatives than any other buyer persona.
This matters for agentic commerce for a specific reason: the more complex and multi-stakeholder the purchase, the more attractive it becomes to deploy an AI agent to manage the coordination burden that includes gathering information, benchmarking options, and preparing recommendations before any human gets involved.
What this Means in Practice
The practical consequence for B2B sellers is architectural. If an AI procurement agent is your buyer, it is not browsing your website the way a human does. It is querying your product data, pricing logic, and contract terms programmatically. It is evaluating structured metadata, not narrative copy. A hallucination in a generated sales memo is an inconvenience; an error made by an autonomous agent operating inside a live procurement system is a commercial and potentially legal event.
Forrester’s research also found that 19% of buyers using AI applications in their purchasing process already feel less confident in their decisions due to inaccurate or unreliable AI-generated information. The implication is not that AI agents will fail, it is that suppliers whose data, pricing structures, and product information are not machine-readable, structured, and verifiable will simply not be considered. The AI agent will route demand elsewhere.
There is a further dimension. Forrester predicts that trials have become a critical risk-reduction strategy in enterprise purchasing, with more than 60% of business buyers now using a trial to evaluate potential solutions, rising to 78% for purchases of $10 million or more. An AI procurement agent evaluating your offering will be looking for exactly this kind of structured, verifiable proof of value. Narrative claims will not clear the threshold.
The Readiness Gap
Gartner has separately predicted that the ungoverned use of generative AI in go-to-market applications will result in the loss of more than $10 billion in enterprise value in 2026 alone, from declining stock prices, legal settlements, and fines. Much of this exposure sits with sellers, not buyers, as AI is integrated into commercial applications without adequate governance.
The businesses that will navigate agentic commerce successfully are those that treat it as a structural question, not a technology one. How is your product information structured? Can your pricing logic be queried programmatically? Does your organisation have governance in place for the AI tools your own go-to-market teams are already using?
The buyer is changing. The sales process built around the human buyer, the relationship, the pitch, the negotiation, is not disappearing. But a growing share of the decision will be made before that conversation ever begins, by systems that do not read whitepapers or attend webinars. Preparing for that reality is no longer a future agenda item.
Source: Gartner, Top Strategic Predictions for 2026 and Beyond; Forrester, The State of Business Buying 2026; Forrester, B2B Marketing, Sales and Product Predictions 2026
